Sports Engineering

https://doi.org/10.1007/s12283-018-0283-0

ORIGINAL ARTICLE

A mathematical model for simulating cycling: applied to track cycling

Billy Fitton1 · Digby Symons1,2

© The Author(s) 2018

Abstract
A review of existing mathematical models for velodrome cycling suggests that cyclists and cycling coaches could benefit
from an improved simulation tool. A continuous mathematical model for cycling has been developed that includes calculated
slip and steering angles and, therefore, allows for resulting variation in rolling resistance. The model focuses on aspects
that are particular, but not unique, to velodrome cycling but could be used for any cycling event. Validation of the model
is provided by power meter, wheel speed and timing data obtained from two different studies and eight different athletes.
The model is shown to predict the lap by lap performance of six elite female athletes to an average accuracy of 0.36% and
the finishing times of two elite athletes competing in a 3-km individual pursuit track cycling event to an average accuracy
of 0.20%. Possible reasons for these errors are presented. The impact of speed on steering input is discussed as an example
application of the model.

Keywords Cycling · Simulation · Velodrome · Dynamics

1 Introduction Olds et al. [1] presented a comprehensive equation for power

demand accounting for aerodynamic drag (including head
1.1 Motivation winds), rolling resistance and equipment specifications.
Olds et al. [2] improved this model, including the impact
Cycling is a sport which lends itself to performance analy- of drafting, crosswinds and the kinetic energy of limbs and
sis. The relative ease of data collection means that com- wheels. This revised model was validated by testing 41 ath-
petitive teams carry out much analysis and spend significant letes over a 26-km time trial. Using measured power as an
resources determining optimal choices of equipment, athlete input the model predicted finishing time to an accuracy of
or strategy. As power meters and other measurements have 5%. Martin et al. [3] developed a model more complicated in
become more accurate, the desire for accurate mathematical its calculation of aerodynamic drag, with wheel drag varying
models has grown. A review of existing predictive cycling with velocity, but otherwise less sophisticated. This model
models has revealed scope for improvement in the modelling had an accuracy of 3% when validated for six test subjects.
of tyre forces in track cycling in particular. Basset et al. [4] derived a model to compare the several dif-
ferent world hour record attempts in the 1990s. This model
1.2 Literature review included some limiting assumptions: i.e. equal groundspeed
and air speed, and frontal surface area, a constant fraction
Since the release of Schoberer Rad Meßtechnik’s (Jülich, of body surface area.
Germany) first power meter in the early 1990s, mathematical It was not until Martin et al. [5] and Lukes et al. [6] who
models of cycling have become increasingly sophisticated. presented models for velodrome cycling that cornering on
a banked track was addressed. Martin et al. [5] considered
the track as circular with constant radius and banking angle.
* Billy Fitton In contrast, Lukes et al. [6] modelled the velodrome as two
[email protected]
straights and two corners, albeit with no transition between
1
Department of Engineering, University of Cambridge, the two. A similar approach was taken by Caddy et al. [7]
Trumpington St, Cambridge CB2 1PZ, UK in an investigation into the impact of cyclist posture on
2
Mechanical Engineering Department, University event performance. All only approximate the true shape of
of Canterbury, Christchurch, New Zealand

Vol.:(0123456789)
 B. Fitton, D. Symons

a velodrome. Lukes et al. [8] refined their approximation by

splitting the velodrome geometry into eight sections rather
than four. This improved model also included tyre scrubbing
effects. It could predict finishing times with an accuracy of +
2%. In a separate study investigating the aerodynamics of
track cycling, Underwood [9, 10] created the most accu-
rate track cycling model to date. Using measured power and
field-derived values for drag area, CdA, the model predicted
elite individual pursuit finishing times with an average error
of 0.42%.
Fig. 1  Summary of forces acting on a cyclist, view is in a plane per-
pendicular to ground surface (β to the vertical)
1.3 Paper overview

Although four of the above models [4, 5, 8–10] concern vel-

odrome cycling, they all neglect some aerodynamic aspects
and model the bends simplistically. To address these limi-
tations, this paper describes a more complete mathemati- ,
cal model for all cycling events. It proposes a method of
predicting the slip and steer angles necessary to navigate
turns (allowing for banking), as well as allowing for rotat-
ing bodies in the system. The model is derived, a method of 1
implementation is explained and then the results of two dif- 2
ferent validation studies are presented. The impact of speed 3
on steering input is discussed as an example application of
the model.

Fig. 2  Forces acting on cyclist in a plane perpendicular to their direc-

2 Mathematical model tion of motion (λ to the vertical)

2.1 Model principles
considered in determining the tyre slip angles (which are of
The focus of this model is velodrome cycling; however, the a similar magnitude).
equations derived are generic and could be applied to other
cycle sports. The model is a significant extension of the 2.2 Derived terms
velodrome-specific model presented in a previous study [7].
The model is a quasi-steady-state analysis that assumes 2.2.1 Governing equation
instantaneous equilibrium of the cyclist at each time step
but allows for changes in speed and configuration between The governing equation for this model is an energy balance:
time steps. At each time step, the cyclist is assumed to be
following a path of known local curvature at constant speed, 𝜂Pin 𝛿t = ΔT + ΔV + Ediss . (1)
i.e. with all accelerations other than centripetal neglected. Over a time period, δt, the available mechanical work
The rate of work done against dissipative forces is calcu- is the product of the cyclist’s input power, Pin, and the effi-
lated based on this instantaneous equilibrium. Any differ- ciency, η, of the bicycle transmission. Drag forces dissipate
ence between the cyclist’s input mechanical work and that much of this energy, Ediss. The remaining power results
dissipated is attributed to changes of gravitational potential in changes in the total kinetic, ΔT, and/or potential, ΔV,
and/or kinetic energy (any change in the latter implying energies.
acceleration). Figure 1 shows forces on the cyclist viewed in a direction
Forces are resolved both tangential and perpendicular to parallel to the ground surface and perpendicular to the direc-
the cyclist’s direction of motion at every time step. The lean tion of motion (see also Fig. 2, a view along the direction
angle, tyre slip angles and steering angle are calculated. The of motion). β is the banking angle of the ground surface;
model assumes that the heading angle of the cyclist, χ, and λ is the angle of vertical inclination of the cyclist’s direc-
the steering input, δ, are small and will, therefore, only be tion of motion and Fλ is the consequent component of the
A mathematical model for simulating cycling: applied to track cycling

cyclist’s weight acting in their direction of motion. Fd is the

aerodynamic drag force and ζ is the angle from the direc-
𝜑=𝜃−𝛽 (10)
tion of motion through which it acts. FR and FN are rolling 2.2.3 Aerodynamic drag
resistance and normal contact forces, respectively, where
subscripts 1 or 2 refer to the front or rear tyre. FF is the Aerodynamic drag can exceed 90% of a cyclist’s resistance to
propulsive force acting at the contact patch of the rear tyre. motion [11, 12]. In this model, the aerodynamic drag force is
Fx is the horizontal component of the centripetal, Fc, and calculated via
drag, Fd, forces (Fig. 2). Fy is the sum of the forces acting
on the cyclist at an angle λ to the vertical and perpendicular | |2
Fd = 1∕2𝜌Cd A|𝐯𝐝∕𝐚𝐢𝐫 | (11)
| |
to the direction of motion (Fig. 2). Both Fx and Fy have
components acting in the plane of Fig. 1 (as shown). Ω is using the drag area, CdA, air density, ρ, velocity of the cen-
the angular velocity of the cyclist and bicycle. tre of drag, vd, and local air velocity, vair. vd/air is the velocity
of the centre of drag relative to the air and is found using
2.2.2 Cyclist dynamics 𝐯𝐝∕𝐚𝐢𝐫 = 𝐯𝐝 − 𝐯𝐚𝐢𝐫 . (12)
Figure 2 shows the forces acting on a cyclist that determine The drag force acts in the same direction as vd/air, at an
the angle of lean, θ. This view is of a plane normal to the angle ζ to the direction of motion and parallel to the ground
direction of motion, i.e. at angle λ to the vertical. Within this surface (it is assumed that vair is in a plane parallel to the
plane, the weight of the cyclist and bike has component Fw. ground surface).
As the wheels navigate a bend with instantaneous radius of
curvature Rw, the cyclist’s centre of gravity moves on a path 2.2.4 Slip, camber and steer angles
with instantaneous radius RCG. The corresponding centrip-
etal force, Fc, acts in a direction perpendicular to both the To determine the necessary steering input, the tyre loading and
direction of motion and the axis of rotation and thus at an slip and camber angles must be calculated (Fig. 3a). The tyre
angle, κ, to the horizontal. Taking moments about the wheel loading depends on the overall equilibrium of the cyclist (see
contact point, and using Figs. 1, 2, 3b). The resultant reaction force, P, acting through
/ the tyres and its components can be derived by
Fc = mvCG 2 RCG , (2)
√ √
Fx = Fc cos 𝜅 + Fd sin 𝜁 cos 𝛽, (3) P= Fx + Fy = FN 2 + FS 2 ,
2 2
(13)

Fw = mg cos 𝜆, (4) FN = P cos 𝜑 = Fx sin 𝛽 + Fy cos 𝛽, (14)

Fy = Fw − Fc sin 𝜅 − Fd sin 𝜁 sin 𝛽, (5) FS = P sin 𝜑 = Fx cos 𝛽 − Fy sin 𝛽, (15)
F𝜆 = mg sin 𝜆, (6) where FN and FS are the total normal contact force and side
( ) force, respectively. FN and FS can be separated into forces
𝜃 = tan−1
Fx acting through front and rear tyres by taking moments about
, (7)
Fy the front tyre contact patch, resolving in two directions and
using the lengths a and b (see Fig. 3b).
Thus it can be found that
RCG = Rw − hCG sin (𝜃 − 𝜅), (8)
( )
an iterative formula for θ can be derived: FN a + Fd cos 𝜁 + F𝜆 hCG cos 𝜑
FN2 = , (16)
a+b
⎛ mvCG 2
cos 𝜅 + Fd sin 𝜁 cos 𝛽 ⎞
−1 ⎜ (Rw − h CG sin (𝜃n −𝜅 )) ⎟
𝜃n+1 = tan ⎜
mvCG 2 ⎟, FN1 = FN − FN2 , (17)
⎜ mg cos (𝜆) − (R − h sin (𝜃 −𝜅 )) sin 𝜅 − Fd sin 𝜁 sin 𝛽 ⎟
⎝ w CG n ⎠ ( )
FS a + Fd cos 𝜁 + F𝜆 hCG sin 𝜑
(9) FS2 = , (18)
where m is the total mass of the cyclist and the bicycle, a+b
g is the gravitational acceleration, vCG is the velocity of
the cyclist/bicycle centre of gravity and hCG the distance FS1 = FS − FS2 . (19)
between the centre of gravity and the wheel/ground contact Having determined the side force required at each tyre,
line. it is possible to calculate the slip and camber angles by the
The roll angle of the cyclist, φ, is given by use of
 B. Fitton, D. Symons

Fig. 3  Top view in a plane - Kinematics

parallel to the ground of a the
kinematics of, and b the forces,
on the bicycle and cyclist
′
2

Ωcos 1

2
1

cos

, cos 1
2 sin

FS F F ( ) 𝛿 cos 𝜀
tan 𝜑 = = S1 = S2 , (20) tan 𝛿 � = . (28)
FN FN1 FN2 cos 𝜑 − 𝛿 sin 𝜑 sin 𝜀

which follows from Eqs. (14)–(18). The ratio of side to nor- Equations (20), (21) and (23) enable the rear tyre slip
mal forces is also given by angle to be given as a function of the known roll angle φ
and the rear tyre stiffness coefficients:
FSi
= 𝛼i C𝛼i + 𝛾i C𝛾i , (21) tan 𝜑 − 𝜑C𝛾2
FNi
𝛼2 = . (29)
C𝛼2
where 𝛼i and 𝛾i are slip and camber angles of the wheel in
question, C𝛼i the cornering stiffness (/rad) and C𝛾i the camber Since α2 is now known Eqs. (26) and (27) can be sub-
stiffness (/rad) of the tyre. stituted into Eqs. (25) and (26) to give
Camber and slip angles for front and rear tyres are defined
by [13] b cos 𝛽∕Rw
𝜒 = sin−1 √ − tan−1 𝛼2 . (30)
𝛼2 2 + 1
sin 𝛾1 = sin 𝜑 + 𝛿 sin 𝜀 cos 𝜑, (22)
𝛾2 = 𝜑, (23) Using Eqs. (20)–(29), it is now possible to obtain a
function of the bicycle and track geometry, tyre coeffi-
𝛼1 = 𝛿 � −
un + aΩ cos 𝛽 cients and roll and heading angles that can be solved itera-
, (24)
ut tively, by substituting Eq. (28) for the steer angle, δ:
( ) ( )
bΩ cos 𝛽 − un C cos 𝛽 C
𝛼2 = , (25) 𝛿 � = tan 𝜒 1 − 𝛼2 + a + b 𝛼2
ut C𝛼1 Rw cos 𝜒 C𝛼1
C𝛾2 C𝛾1 −1 (31)
where ε is the steering rake angle. un and ut are the nor- + 𝜑− sin (sin 𝜑 + 𝛿 sin 𝜀 cos 𝜑).
C𝛼1 C𝛼1
mal and tangential components of the bicycle velocity with
respect to the bicycle frame, see Fig. 3a and 2.2.5 Rolling resistance
un = vw sin 𝜒, (26)
The total rolling resistance is given by
ut = vw cos 𝜒, (27)
FR = FR1 + FR2 = FN1 Crr1 + FN2 Crr2 , (32)
where χ is the heading angle. The relationship between
ground steer angle, δ′ and steer angle, δ, is given by [13] where Crr1 and Crr2 are coefficients of rolling resistance for
front and rear tyre, respectively. These coefficients depend
A mathematical model for simulating cycling: applied to track cycling

[ v ]
𝛚𝐰∕𝐜 = − rw 0 0 , (35)

where r is the outer radius of the tyre and the components

are in the first, second and third directions (Fig. 2).
( − ) The cyclist and bicycle frames rotate as a rigid body,
with an angular velocity of
vc v v v
Ω= = CG = cw = w . (36)
Rc RCG Rcw Rw

The components of the wheels’ and cyclist’s angular

velocities in the first, second and third directions are given
by

Fig. 4  Showing velocities of the cyclist and bicycle in a plane parallel

𝛚𝐜 = 𝛀 = [Ω sin(𝜃 − 𝜅) Ω sin(𝜃 − 𝜅) 0], (37)
to the bicycle frame 𝛚𝐰 = 𝛚𝐰∕𝐜 + 𝛚𝐜 , (38)
[ v ]
on the characteristics of the particular tyre and the instanta- 𝛚𝐰 = Ω sin(𝜃 − 𝜅) − w Ω cos(𝜃 − 𝜅) 0 . (39)
neous slip and camber angles. Measurement of the rolling r
resistance characteristics of a tyre requires careful experi-
mentation, see, e.g. Fitton and Symons [14]. Due to the quasi-steady-state approach adopted for this
model, the component of angular velocity in the third
2.2.6 Potential energy direction (i.e. dθ/dt, the rate of change of lean angle) is
assumed to be zero at each instant.
Work done against gravity is determined by changes in Moments of inertia about the centre of gravity in the
potential energy, V, which is equal to first, second and third directions (Fig. 2) are assumed to
( ) be principal moments of inertia for both wheel and cyclist;
V = mgz = mg hw + hCG cos 𝜃 , (33) thus:
where the overall height of the centre of gravity, z, depends
on both the varying height of the path of the wheels, hw, and ⎡ Iw1 0 0 ⎤ ⎡ Ic1 0 0 ⎤
the lean angle of the cyclist, θ. 𝐈𝐰 = ⎢ 0 Iw2 0 ⎥ 𝐈𝐜 = ⎢ 0 Ic2 0 ⎥. (40)
⎢ ⎥ ⎢ ⎥
⎣ 0 0 Iw3 ⎦ ⎣ 0 0 Ic3 ⎦
2.2.7 Kinetic energy
With all terms in Eq. (34) determined, the total kinetic
The kinetic energy of the system is given by energy, T, can be defined in terms of vCG by
( ) ( )
1 1 1 1
T = 2 mw 𝐯𝐜𝐰 2 + mc 𝐯𝐜 2 + 2 𝐈𝐰 𝛚𝐰 2 + 𝐈𝐜 𝛚𝐜 2 ,
T = KvCG 2 , (41)
2 2 2 2 where
(34)
( ( ) ( ) ( )2
where mw and Iw are the mass and moment of inertia of one Rcw 2 1 Rc 2 sin (𝜓) R
bicycle wheel, vcw and ωw are the translational and angular T= mw + mc + Iw1 − w
RCG 2 RCG RCG RCG r
velocities of the wheel, respectively, mc and Ic are the mass ( )2 ( (
) cos (𝜓) ) 2
) (42)
and moment of inertia of the cyclist/bicycle, and vc and ωc 1
+ Ic1
sin (𝜓) 1
+ Iw2 + Ic2 vCG 2
2 RCG 2 RCG
are the translational and angular velocities of the cyclist/
bicycle. Rotational kinetic energy of the limbs, pedals and and
cranks is neglected; Olds [2] showed them to account for
only 0.07% of total kinetic energy, compared to 2% for the
𝜓 = 𝜃 − 𝜅. (43)
wheels. To make use of Eq. (34), the different velocities
must be determined from the geometry of the cyclist and 2.3 Numerical solution and implementation
bicycle (Fig. 4).
By assuming no longitudinal slip for the tyre/ground con- A straightforward implementation is forward integra-
tact, the angular velocity of the wheel relative to the cyclist, tion of the acceleration, aCG, over fixed time increments,
ωw/c, can be found by δt. Forces and configuration are calculated assuming
 B. Fitton, D. Symons

instantaneous steady-state cornering using the method 3 Validation

above; both are assumed constant throughout the time step.
aCG is calculated, and also assumed constant throughout 3.1 Method, assumptions and fixed terms
the time step. This approach is a computationally efficient
approximation that should be sufficiently accurate if δt 3.1.1 Method
remains small (i.e. less than 2 s).
The rate at which energy is lost to dissipative forces is The model was implemented in Matlab (Mathworks, Cam-
the product of the force magnitudes and the corresponding bridge, UK) and validated by two different methods. The
velocities and can, therefore, be calculated by investigation has been approved by the Cambridge Univer-
dEdiss sity Engineering Department Research Ethics Committee.
= Fd cos 𝜁vd + FR vw . (44) First, the tool was used to predict the lap times of six dif-
dt
ferent elite female athletes cycling at approximately constant
Differentiating Eq. (33) means that the rate of change of speed. These athletes each took part in three sub maximal
potential energy can be determined by efforts at the Manchester velodrome on three separate days.
Throughout each effort, the athletes were asked to maintain
( )
dV d𝜃 a specified speed and to follow the 250-m datum line as
= mg vw sin 𝜆 − hCG sin 𝜃 . (45)
dt dt closely as possible. The specified speed was varied for each
session and athlete. In total, the six athletes completed 174
Note that the quasi-steady-state approximation assumes laps at speeds between 42 and 51 km/h. The recorded power
that the angular velocity of roll, dθ/dt, is approximately zero data and measured athlete characteristics were then used to
at each time step; therefore, in the simplest implementation predict the athlete’s performance throughout the 174 laps.
dθ/dt must be estimated via linear extrapolation from previ- All of the participants gave informed consent for their data
ous time steps. to be used in this investigation.
If the input power Pin of the cyclist is known, the power Second, the tool was used to predict the finishing time
PT associated with a change in kinetic energy can be calcu- of two elite female cyclists competing in the 3 km Individ-
lated by modifying Eq. (1) to ual Pursuit (3KIP) event at the 2017 UEC European Track
dT dV dEdiss Championships (ETC2017) in Berlin from the input power
PT = = 𝜂Pin − − . (46) recorded for each cyclist during the same event.
dt dt dt
If K is assumed to be approximately constant over a time 3.1.2 Athlete power
step then PT can be calculated by
d ( 2) Input power of the athletes was recorded using a power
PT ≈ K v = 2KvCG aCG . (47) meter (Schoberer Rad Meßtechnik, Germany) which had
dt CG
been calibrated according to the manufacturer’s instructions.
The acceleration of the cyclist’s centre of gravity is given
by 3.1.3 Atmospheric conditions
PT
aCG = . (48) Air density, ρ, was calculated from local atmospheric con-
2KvCG ditions at the time using Teten’s formulation [15]. Gravi-
The velocity, displacement, configuration and forces act- tational acceleration, g, was determined for the velodrome
ing on the cyclist at the beginning of the next time step can locations [16].
then be determined.
To avoid the discontinuity that arises from vCG equal- 3.1.4 Track geometry and cyclist trajectory
ling zero at the start of the initial time step, Eq. (48) can be
modified to Track geometry (banking angles, radii, inclinations) was
required for two different velodromes. The Manchester (UK)
Rcw velodrome geometry was found from a survey of the 250-m
aCG = Q , (49)
KRCG rG datum line using a TC403L total station (Leica Geosystems,
Heerbrugg, Switzerland). The geometry of the Berlin velo-
where G is the gear ratio and Q the starting torque. drome was determined from a combination of expert knowl-
edge and information given by the track designers: Schuer-
mann Architects (Muenster, Germany). It was assumed that
A mathematical model for simulating cycling: applied to track cycling

the wheels of the cyclists exactly followed the datum line. 3.1.7 Efficiency of the bicycle
In contrast to other studies [4, 5, 7, 8] the altitude, hw, of the
datum line was allowed to vary. Sources of inefficiency on a bicycle include drivetrain, frame
flexibility and wheel bearings. For this validation, a fixed
3.1.5 Drag area and aerodynamic drag mechanical efficiency, η, of 98% [19] has been assumed.

Bulk airflow is caused by cyclists circling a velodrome. 3.1.8 Tyre properties

Throughout the validation process, this airflow was assumed
to remain constant in both magnitude and direction. The Tyre properties Cα, Cγ and Crr were determined in a previ-
magnitude of the airflow was the average of that meas- ous study [14]. Cα and Cγ were found to be constant and
ured during the validation session using an anemometer. independent of tyre normal force, whereas Crr was found
The direction of the airflow was assumed tangential to the to be a function of both the loading and orientation (α and
cyclist’s motion, i.e. ζ = 0. γ) of the tyre. Crr has, therefore, been modelled to increase
Due to lack of access to a wind tunnel, the cyclist’s drag non-linearly with both α and γ, and most significantly with
area CdA was derived from field testing. Each athlete under- the former.
went aerodynamic testing with identical equipment and
maintaining the same position as that used in each effort
but at an earlier date. The protocol outlined by Fitton et al. 4 Results
[17] was used in each instance. CdA was assumed constant
throughout each effort. For part one of this validation, lap times were predicted with
Note that the centre of drag (the point through which the an average accuracy of 0.36%. The maximum and minimum
aerodynamic drag force acts), is assumed the same as the errors of the model were 0.98% and 0.001%, respectively.
centre of gravity. The cyclists’ positions, and the fact that The standard deviation of the errors was 0.22%. The predic-
their bodies typically account for 70% of their total aerody- tion was less than the actual lap time in 86% of cases; a prob-
namic drag [11], means that the centre of drag is very close able explanation is the cyclists’ imperfect handling ability.
to the centre of gravity. Analysis of the measured wheel speed data revealed that the
elite cyclists travelled 0.7% further than the track length.
3.1.6 Mass and inertia In the second validation study, finishing times of the two
events were predicted with an average accuracy of 0.20%
In the first part of this validation, the mass of the cyclists and individual split times with an average absolute accu-
and their equipment was measured before and after each racy of 0.24% (Table 1). Figure 5 compares simulated and
session and the average used in the simulation. In the second recorded wheel speed data for Athlete A. Again the simu-
part mass was measured only once, as close to the event as lation under predicted the finishing time of both athletes.
practical. An additional contributing factor may be the assumption of
The model requires the centre of gravity location and constant CdA throughout the event. In a 3KIP CdA may be
moment of inertia of the cyclist. These inputs were deter- greater at the start, as the athlete pedals out of the saddle,
mined by measuring an average-sized elite cyclist, who did and at the end of the event, where the tiring cyclist may not
not take part in either study but was part of the same team, maintain their position.
from a high-definition photo and then modelling each limb, The error associated with the simulation’s prediction of
the torso and the head as separate ellipsoids. The same finishing times and, importantly, split times is lower than
cyclist was weighed and each ellipsoid assigned a proportion in any previous comparable study. The model described by
of the cyclist’s total mass typical for an average human [18]. Lukes et al. [8] was capable of predicting finishing time in
Using two more photos of the same athlete in their cycling a 4KIP to within 2% and individual split times (0–1 km,
position, three-dimensional coordinates were assigned to 1–2 km, 2–3 km, 3–4 km) with a slightly larger error than
each limb. With the limb dimensions, estimated masses and that. Underwood’s [9, 10] proposed model was able to pre-
positions, it was possible to determine the centre of mass and dict finishing times for elite athletes competing in the 3KIP
the inertia of the cyclist and bicycle (without wheels), IC. and 4KIP to within 0.42%. When investigating Underwood’s
The values for other athletes were determined by scaling for model’s prediction of split times (0–1 km, 1–2 km, 2–3 km),
their relative physical characteristics. Wheels were assumed however, the available data suggest higher errors of approxi-
to be uniform discs to calculate inertia. mately 2.5%.
Significant factors contributing to the accuracy of the
model presented here include the consideration of rotational
kinetic energy and varying tyre forces and the care taken in
 B. Fitton, D. Symons

Table 1  Comparison of actual [20] and simulated split times for the 3KIP events at the ETC2017
Split Actual Simulated Total error (%) Split error (%)
Total time (s) Split time (s) Total time (s) Split time (s)

Athlete A
0–1000 m 73.196 73.196 73.049 73.049 − 0.20 − 0.20
1000–2000 m 140.467 67.271 140.265 67.215 − 0.14 − 0.08
2000–30,000 m 209.328 68.861 208.788 68.523 − 0.26 − 0.49
Athlete B
0–1000 m 74.313 74.313 73.992 73.992 − 0.43 − 0.43
1000–2000 m 146.726 72.413 146.312 72.320 − 0.28 − 0.13
2000–30,000 m 223.157 76.431 222.833 76.521 − 0.15 0.12

Fig. 5  Comparison of simulated

and recorded wheel speed for
cyclist competing in the 3KIP at
the 2017 ETC2017

measuring inputs, particularly coefficients of aerodynamic 6 Conclusions

drag and rolling resistance.
A mathematical model of simulating cycling has been
developed. The model includes aspects of particular rel-
5 Example application evance to velodrome cycling. Via two different validation
studies, the accuracy of the model has been shown to sur-
One novel aspect of this model is the capability to predict pass previous comparable models: errors in predicted lap
tyre slip angles and the necessary steering input, δ, to navi- times are consistently less than 0.36%. A key advantage of
gate a particular trajectory. Using the geometry of the Man- the model is the calculation of steer and tyre slip angles;
chester velodrome datum line, the impact of speed on δ at this enables the rolling resistance to be predicted more
the bend apex has been predicted for Athlete A (Fig. 6). accurately. This makes it possible, for example, to com-
At low speeds, the model predicts a low δ despite the low ment on the impact of handling ability and tyre choice on
lean angle, θ (Fig. 7). As speed increases δ also increases to event performance.
a peak of ~ 1.7° at ~ 50 km/h. This approximately coincides
with the speed at which roll angle, φ, equals zero. As speed
and φ further increase δ is predicted to decrease.
A mathematical model for simulating cycling: applied to track cycling

Fig. 6  The impact of speed on

the necessary steering input
at the apex of the bend in the
Manchester Velodrome

Fig. 7  The impact of speed on

the cyclist’s lean and roll angle
at the apex of the bend in the
Manchester Velodrome

Acknowledgements The authors would like to thank the study par- 3. Martin JC, Milliken DL, Cobb JE et al (1998) Validation of a
ticipants and staff at British Cycling, EIS and Manchester Velodrome. mathematical model for road cycling power. J Appl Biomech
14:276–291
Open Access This article is distributed under the terms of the Crea- 4. Bassett DR, Kyle CR, Passfield L et al (1999) Comparing cycling
tive Commons Attribution 4.0 International License (http://creat​iveco​ world hour records, 1967–1996: modeling with empirical data.
mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distribu- Med Sci Sports Exerc 31:1665–1676
tion, and reproduction in any medium, provided you give appropriate 5. Martin JC, Gardner AS, Barras M, Martin DT (2006) Mod-
credit to the original author(s) and the source, provide a link to the eling sprint cycling using field-derived parameters and forward
Creative Commons license, and indicate if changes were made. integration. Med Sci Sports Exerc 38:592–597. https​: //doi.
org/10.1249/01.mss.00001​93560​.34022​.04
6. Lukes R, Carré M, Haake S (2006) Track cycling: an analytical
model. In: The engineering of sport 6, vol 1: developments for
sports. Springer, New York, pp 115–120
References 7. Caddy O, Fitton W, Symons D, Purnell A, Gordon D (2015) The
effects of forward rotation of posture on computer-simulated
1. Olds T, Norton KI, Craig NP (1993) Mathematical model of 4-km track cycling: implications of Union Cycliste Internation-
cycling performance. J Appl Physiol 75:730–737 ale rule 1.3.013. Proceedings of the Institution of Mechanical
2. Olds T, Norton KI, Lowe EL et al (1995) Modeling performance. Engineers, Part P: Journal of Sports Engineering and Technology
J Appl Physiol 78:1596–1611 231(1):3–13
 B. Fitton, D. Symons

8. Lukes R, Hart J, Haake S (2012) An analytical model for Sport VII: 8th Asia-Pacific Congress on Sports Technology. Tel
track cycling. Proceedings of the Institution of Mechani- Aviv, Israel, 15–19 October 2017
cal Engineers, Part P: Journal of Sports Engineering and 15. Tetens O (1930) Über einige meteorologische begrifle. Z Geophys
Technology 226(2):143–151 6:297–309
9. Underwood L (2012) Aerodynamics of track cycling. PhD Thesis, 16. Li X, GÓ§tze H-J (2001) Tutorial ellipsoid, geoid, gravity, geod-
Univ Canterbury, New Zealand esy, and geophysics. Geophysics 66:1660–1668
10. Underwood L, Jermy M (2014) Determining optimal pacing strat- 17. Fitton B, Caddy O, Symons D (2017) The impact of relative ath-
egy for the track cycling individual pursuit event with a fixed lete characteristics on the drag reductions caused by drafting when
energy mathematical model. Sport Eng. https​://doi.org/10.1007/ cycling in a velodrome. Proc Inst Mech Eng Part P J Sport Eng
s1228​3-014-0153-3 Technol. https​://doi.org/10.1177/17543​37117​69228​0
11. Kyle CR, Burke ER (1984) Improving the racing bicycle. Mech 18. MacDougall JD, Wenger HA, Green HJ (1991) Physiological
Eng 106:34–45 testing of the high-performance athlete. Human Kinetics Books,
12. Grappe F, Candau RB, Belli A, Rouillon JD (1997) Aerodynamic Champaign
drag in field cycling with special reference to the Obree’s position. 19. Kyle CR, Burke ER (2003) Selecting cycling equipment. High
Ergonomics 40:1299–1311 Tech Cycl 2:1–48
13. Pacejka H (2005) Tire and vehicle dynamics. Elsevier, Amsterdam 20. Union Européenne De Cyclisme (2017) 2017 UEC track elite
14. Fitton B (October 2017) Symons D (2017) The characterisation of european championships results book. Union Européenne De
bicycle tyres and their impact on event performance for athletes Cyclisme resources. http://uec.ch/resou​rces/PDF/2017%20Tra​
of differing abilities, presented to The Impact of Technology on ck%20Ber​lin/resul​ts/Resul​ts%20Boo​k.pdf